![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
(no subject)
Sep. 21st, 2007 10:56 pmAaaand today I bounced off another friend at utter random in a bar in Holborn, which makes two in two days.
However I have no grounds for suprise. An a priori estimate is not really useful - the reasonable range for parameters going in (which include number or people I can recognise in a crowd from the shape of a shoulder or a shift of weight, number of people I pass by close enough to notice in any one day, relevant population in which the people I might recognise are scattered) leads to an range which covers about 4 orders of magnitude. So I fall back to the only point estimate I have: in about 660-700 days of working London I've met only about 7 people at random in the street: that's 1/100 per day, or 1/10,000 for two in two days as independent events.
And 1/10,000 does not justify suprise - if the relevant population is say 10 million then it happens a thousand times a day, and the weak inductive principle (or Bayes' theorem, whichever way you prefer to conceptualise it) means that I cannot be suprised just because it happened to me.
I can definitely be startled as all hell, though. Still, it was nice to see J., and it's a useful reminder that naive intuition is very nearly worthless.
However I have no grounds for suprise. An a priori estimate is not really useful - the reasonable range for parameters going in (which include number or people I can recognise in a crowd from the shape of a shoulder or a shift of weight, number of people I pass by close enough to notice in any one day, relevant population in which the people I might recognise are scattered) leads to an range which covers about 4 orders of magnitude. So I fall back to the only point estimate I have: in about 660-700 days of working London I've met only about 7 people at random in the street: that's 1/100 per day, or 1/10,000 for two in two days as independent events.
And 1/10,000 does not justify suprise - if the relevant population is say 10 million then it happens a thousand times a day, and the weak inductive principle (or Bayes' theorem, whichever way you prefer to conceptualise it) means that I cannot be suprised just because it happened to me.
I can definitely be startled as all hell, though. Still, it was nice to see J., and it's a useful reminder that naive intuition is very nearly worthless.
